[Original Post that started the Blood Analysis thread, around 7 Nov 1990]

From: Walter Henry ( whenry@lindy.stanford.edu )

>A colleague at another library has asked what I think is a rather
>interesting question.  Her institution has a manuscript that was
>ostensibly signed in blood in the 17th century and we are trying to
>think of a way to verify whether this is the case.  Does anyone know of
>a test that would identify blood (or its components) that
>        a) requires only microsamples?
>        b) might work on a sample that is 300 years old?
>Any suggestion, or pointers to literature would be most welcome.  
>I do recall something of the sort in a rather noxious popular book on
>the Shroud of Turin and will try to find that, and I dare say the
>Forensic literature is rife with such matters, but alas I don't have
>much access to such arcana.
>Thanks,
>Walter Henry

....

 Article 2451 of sci.chem:
 From: John G. DeArmond <jgd@rsiatl.UUCP>
 Date: 11 Nov 90
 Organization: Radiation Systems, Inc. (a thinktank, motorcycle, car
     and gun works facility)
 ...

 You might want to look at X-ray Fluorescence Analysis which works by
 stimulating the unknown with X-rays and analysing the fluorescence that
 results.  With the proper equipment and software, it is possible to
 "strip away" layers of thin film, such as on hard disk platters and
 make chemical analyses to 2 decimal point resolution.  Low Z materials
 such as the constituents of paper do not show up and/or easily
 filtered. We just finished a project that involved building some
 portable instruments for the Smithsonian.  You could contact them or I
 could put you in touch with my colleague who is a recognized expert in
 the field.

 John
 --
 John De Armond, WD4OQC        | "Purveyors of Performance Products
 Rapid Deployment System, Inc. |  to the Trade " (tm)
 Marietta, Ga                  |
 {emory,uunet}!rsiatl!jgd      | "Vote early, Vote often"


 Article 1276 of sci.chem:
 From: Larry Lippman <larry@kitty.UUCP>
 Summary: Limitations of x-ray fluorescence
 Date: 24 Nov 90
 Organization: Recognition Research Corp., Clarence, NY
 In article ..., jgd@rsiatl.UUCP (John G. DeArmond) writes:

 > You might want to look at X-ray Fluorescence Analysis which works by

     XRF (x-ray fluorescence) is a technique used for elemental analysis
 above a Z (atomic number) of 8 or so.  Therefore, it cannot readily
 detect the common elemental constituents of organic materials such as
 carbon, hydrogen, oxygen and nitrogen.  Special XRF apparatus using a
 vacuum chamber has been able to detect elements with a Z as low as 5
 (boron), but this is not a particularly sensitive, convenient or
 reliable technique.

     While it is obvious that the heme molecule contains iron, XRF
 cannot readily discriminate between heme and say, an ink containing
 just iron oxide (like red ocher).  While XRF may be able to
 affirmatively identify certain inks with compositions comprising
 multiple elements above atomic number 8, it cannot affirm or deny the
 presence of blood with any reasonable degree of certainty.  A good
 example of XRF for ink identification is the inks used to print
 currency, which have far more elements that one might imagine; a
 typical dollar bill contains: titanium, barium, lead, chromium, iron,
 cobalt, zinc and tungsten!

     XRF is quite useful for identification and authentication of many
 inks and pigments in documents, paintings, and other archaeological
 objects. But it just won't do blood.

     Now that we have dispensed with XRF, there *is* another x-ray
 method which can affirmatively identify blood - but it's not very easy.
 XRD (x-ray diffraction) can identify the specific structure of the heme
 molecule, and was in fact used by John Kendrew during the 1950's to
 discover the precise structure of hemoglobin.

     However, sample preparation for XRD requires some destruction of
 the article under investigation.  In addition, it is necessary to
 separate the ink (or blood) from the paper and any other materials to
 minimize the number of unknowns.  XRD can usually deal with binary
 mixtures, but beyond that it become extremely complex, if not
 impossible.  The sample, possibly mixed with a binder, is placed in a
 capillary tube or molded, and then inserted within the XRD camera.

     While XRD may be used to affirmatively identify heme, it is not a
 simple process.  Some considerable computer-aided analysis of XRD
 patterns is required.  The complete structure of hemoglobin results in
 XRD patterns with over 25,000 reflections!

 > We just finished a project that involved building some portable
 > instruments for the Smithsonian.

     Interesting.  I know some people at both the Freer Gallery and the
 Museum Support Center in Suitland, MD.  I'll have to ask them about
 your project.

 Larry Lippman @ Recognition Research Corp.  "Have you hugged your cat today?"
 VOICE: 716/688-1231   {boulder, rutgers, watmath}!ub!kitty!larry
 FAX:   716/741-9635                  {utzoo, uunet}!/      \aerion!larry



 Article 1277 of sci.chem:
 From: Larry Lippman <larry@kitty.UUCP>
 Summary: ARCHAEOLOGICAL dating by amino acid racemization measurement
 Date: 24 Nov 90
 Organization: Recognition Research Corp., Clarence, NY

 In article <...>;, mikew@sanjuan.wrcr.unr.edu (Mike Whitbeck) writes:
 > I find this interesting! I have had casual inquiries from
 > ARCHAEOLOGISTS about this. As far as I know the simple forensic tests
 > (field tests) simply check for iron-- not terribly specific.

     Not iron, really, but many of the simple tests involve peroxidase
 reactions of hemoglobin (as I have described in a recent article).  The
 only test which particularly reacts to iron is luminol - this test is
 seldom used anymore due to its propensity for interference from other
 metallic sources.

 > The question that must be answered is 'Can blood proteins, in whole or
 > sufficiently unique fragments survive over 300 years?'

     Sure!  In fact, organic materials can often be dated by measurement
 of the extent of racemization of certain amino acids in its constituent
 proteins.  An example of such a racemization reaction having a
 half-life of approximately 100,000 years is: L-isoleucine --
 D-alloisoleucine. Another racemization mechanism is L-aspartic acid
 -- D-aspartic acid, which has a half-life of approximately 17,500
 years.

     So, clearly if we can date proteins to 100,000 or more years, we
 can certainly take other measurements using 300-year old blood.

 >If the answer is yes then by all means talk to a
 >biochemist about electrophoresis or chromatographic methods.

 Unfortunately, there are some limitations.  While we can identify
 certain proteins and such substances as hemoglobin (heme, actually),
 many of the characteristic enzymes, antigens and globulins in blood are
 long destroyed by age.  However, sufficient fragments exist such that
 amino acids and other components may still be identified.

 Larry Lippman @ Recognition Research Corp.
 VOICE: 716/688-1231   {boulder, rutgers, watmath}!ub!kitty!larry
 FAX:   716/741-9635 {utzoo, uunet}!/\aerion!larry



 Article 1291 of sci.chem:
 From: Larry Lippman <larry@kitty.UUCP>
 Summary: Identification of blood stains
 Date: 24 Nov 90
 Organization: Recognition Research Corp., Clarence, NY

 In article <...> tjbryce@amherst.bitnet writes:
 ...
 >You might consider testing the wavelength at which the "blood" absorbs
 >light. Hemoglobin absorbs light at very specific frequencies, which would
 >be a useful fingerprint to tell the "blood" apart from ink.

     I agree with the above advice.  A microspectrophotomet is an
 instrument which couples a traditional scanning UV/VIS/NIR
 spectrophotometer to a microscope, and can facilitate both transmissive
 and reflective measurements.  In a viewing mode the microscope eyepiece
 is used to position the instrument to the desired location on the
 sample, at which point a movable mirror is operated to connect the
 spectrophotometer in place of the eyepiece.  Since one must use a more
 complex optical path through a microscope, the wavelength of a
 microspectrophotomet is limited to not much more than 1100 nm at the
 high end, and not much less than 300 nm at the low end.

     Zeiss makes a series of microscopes called "Zonax" which offer
 scanning @spectrophotometric capabilities.  A computer with color
 monitor is used to process and display various types of
 spectrophotometric data representations.  My organization had a Zeiss
 Zonax system for almost two years as GFE; unfortunately, we had to
 return it at the conclusion of the contract. :-)  This particular model
 was really slick, although its spectral range was limited to 400 to 700
 nm.  Newer models have greater spectral range.  Zeiss Zonax systems are
 rather pricey, however; the model we had sold for around $ 60 K.

     Getting back to the specific issue of blood, hemoglobin has two
 very specific absorption bands at 540 and 578 nm, with a minimum
 absorption at 560 nm.  Not only can such spectrophotometric data be
 used for determination of the presence of blood, but work by Kind,
 Patterson et al has developed methodology for the spectrophotometric
 *dating* of blood. In this case, a small amount of a bloodstain is
 dissolved in a dilute solution of ammonium hydroxide, and a complete
 spectrophotometric scan is made.  Analysis of the spectrophotometric
 curve is used to arrive at an approximate age of the bloodstain.
 Further details on this process is in the literature.

     In the particular case of authenticating the document in question,
 I would suggest that a spectrophotometric determination be made along
 with at least one *other* test.  The following are general suggestions
 for such an additional test (explicit test details are readily found in
 the literature and in textbooks on forensic science).  Sampling for
 these tests can be relatively non-destructive through scraping of a
 small sample of suspected blood, or through the use of moistened filter
 paper to absorb a sample of the suspected blood.

 1.  Methylated benzidine test, which can detect blood in concentrations
     as low as 10 ppm.

 2.  Kastle-Meyer phenolphthalein test, which while less sensitive than
     the methylated benzidine test, is surprisingly immune to
     interference (except for a few fruit and vegetable materials).

 3.  Takayama hemochromogen test, which uses pyridine to cause the
     reduction of hemoglobin, resulting in characteristic salmon-pink
     crystals of pyridine hemoglobin observable under the microscope.

     Unfortunately, at this point, even if we achieve affirmative
 results from the above tests, we are faced with another problem: is the
 blood *human*?

     With blood stains that are less than a few months old, this is not
 usually a problem, and various procedures exist for such
 determinations. A common test uses what is called anti-human
 precipitating antiserum. There is, however, a common interference
 problem with such precipitin tests resulting from tannin in wood, paper
 and leather.  This may or may not be a problem with the document in
 question.  The use of anti-serum tagged with fluorescent dye, and the
 making of observations under a microscope with UV illumination may aid
 in precipitin testing.  Another problem, though, is that anti-human
 serum will not differentiate between human and primate blood - however,
 I would probably not lose any sleep over this remote possibility. :-)

     There is a gel electrophoresis procedure whereby the albumin and
 non-gamma globulins of the samples are made to migrate toward the gamma
 globulins of the selected anti-sera (usually a selection of anti-human,
 anti-cow, anti-pig, anti-dog, etc.).  Positive results are indicated by
 a precipitation line.

     Unfortunately, it is highly doubtful that 300-year old blood stains
 can be species-identified (let alone ABO-typed) by *any* method. Many
 antigens, such as M and N factors are destroyed within several weeks.
 Various isoenzymes and serum proteins (like Hp, AK, PGM, EAP, etc.) are
 also destroyed within a few months.  I have, however, heard it claimed
 that if blood dries quickly and *remains* dry, the A and B antigens
 will remain stable almost indefinitely.

     The oldest antigen identification that I have heard of occurred in
 the U.K. after a period of ten years.  This specimen actually contained
 a dried crust of blood, as opposed to a stain - so the greater quantity
 may have been a factor.

     While I feel confident that 300-year old blood can be reliably
 identified *as blood*, I am highly doubtful that any species
 identification can be made.  I don't know if that is sufficient
 authentication for the document in question.

 Larry Lippman



 Article 1292 of sci.chem:
 From: Roy Smith < roy@phri.nyu.edu >
 Date: 25 Nov 90
 Organization: Public Health Research Institute, New York City

 larry@kitty.UUCP (Larry Lippman) writes:
 > I feel confident that 300-year old blood can be reliably identified
 > *as blood*, I am highly doubtful that any species identification can
 > be made.

     Would it be possible to recover enough DNA to do PCR on it?  I seem
 to remember reading something about doing PCR on DNA from a frozen
 quagga (a zebra-like beast which became extinct in the last ice age,
 now only found on "rogue" games).  The quagga DNA was O(10,000) years
 old, which is considerably older than 300 years, but perhaps freezing
 is a better method of preserving than drying on paper?

     Assuming you could do PCR, how hard would it be to determine the
 species?  The most likely other candidates would probably be some sort
 of food animal, such as cow, pig, or chicken (or whatever animals they
 were eating 300 years ago; the point is, other primates would be
 unlikely). Actually, it just occurred to me that red blood cells don't
 have DNA, if I remember correctly, but maybe some other component of
 blood does?
 ...
 --
 Roy Smith, Public Health Research Institute
 455 First Avenue, New York, NY 10016
 roy@alanine.phri.nyu.edu -OR- {att,cmcl2,rutgers,hombre}!phri!roy
 "Arcane?  Did you say arcane?  It wouldn't be Unix if it wasn't
 arcane!"



  Article 1295 of sci.chem:
 From: Jeff Forbes &lt;forbes@aries.scs.uiuc.edu&gt;
 Date: 25 Nov 90
 Organization: School of Chemical Sciences,
               Univ. of Illinois at Urbana-Champaign

 In article <...> roy@phri.nyu.edu (Roy Smith) writes:

 > Actually, it just occurred to me that red blood cells don't have DNA,
 > if I remember correctly, but maybe some other component of blood does?

 You are correct. Mammalian erythrocytes do not have a nucleus and
 therefore no DNA. Other cellular components of blood do have a nucleus,
 but the concentration is much less than the erythrocytes. Avian
 erythrocytes do have a nucleus.

 Jeff Forbes

 "....I have not failed. I've just found 10,000 ways that won't work."
             Thomas Edison



 Article 1298 of sci.chem:
 From: Andrew Taylor ( andrewt@cs.su.oz )
 Date: 25 Nov 90
 Reply-To: andrewt@cluster.cs.su.oz (Andrew Taylor)
 Organization: Basser Dept of Computer Science
               University of Sydney, Australia
 ...
 Quaggas became extinct about 100 years ago (in Southern Africa). Maybe
 you have the wrong animal?



 Article 1300 of sci.chem:
 From: Mike Whitbeck < mikew@sanjuan.wrcr.unr.edu >
 Date: 26 Nov 90
 Reply-To: mikew@sanjuan.UUCP (Mike Whitbeck)
 Organization: DRI-WRC Reno

 I recently read an article in the newspaper that some
 archaeologists were using serum antigens to ID very old blood
 with species identification.
 ______________________________.
| mikew@wheeler.wrc.unr.edu    |
|__RENO___NEVADA_______________|



 Article 1302 of sci.chem:
 From: Larry Lippman &lt;larry@kitty.UUCP&gt;
 Summary: Unlikelihood of DNA sequencing from old, dried blood
 Date: 26 Nov 90
 Organization: Recognition Research Corp., Clarence, NY

 In article <...> roy@phri.nyu.edu (Roy Smith) writes:
 > larry@kitty.UUCP (Larry Lippman) writes:
 > Would it be possible to recover enough DNA to do PCR on it?

     I don't believe that any DNA fragments would exist in the dry
 condition of a 300-year old blood stain, notwithstanding the fact that
 mature mammalian erythrocytes don't have a nucleus, mitochondria,
 ribosomes, etc.  Leucocytes are a true cell with a nucleus and DNA, but
 I rather doubt that any would be found in a 300-year old blood stain.

     I can't offer any authoritative comments on DNA issues, though. I
 don't do DNA, genetic mapping or any aspect of molecular biology. I
 have enough trouble maintaining competency in selected more traditional
 aspects of chemistry and biochemistry! :-)

 > Actually, it just occurred to me that red blood cells don't have DNA,
 > if I remember correctly, but maybe some other component of blood does?

     As far as I can recall, no *mature* mammalian erythrocytes possess
 DNA.  Erythrocytes of birds, reptiles, etc. do have a nucleus with DNA
 and all the trimmings, though. ...

 Larry Lippman



 Article 1310 of sci.chem:
 From: Eric E. Snyder < eesnyder@boulder.Colorado.EDU >
 Date: 27 Nov 90
 Organization: University of Colorado, Boulder
 ...
 In article <...> larry@kitty.UUCP (Larry Lippman) writes:
 >In article <...> roy@phri.nyu.edu (Roy Smith) writes:
 >>  Would it be possible to recover enough DNA to do PCR on it?
 > I don't believe that any DNA fragments would exist in the dry
 > condition of a 300-year old blood stain.....

 I bet there would be PCRable DNA fragments....  Remember that DNA is a
 remarkably stable molecule (even in my hands).  Furthermore, PCR only
 requires a few molecules of template to give a signal... that is not
 too much to ask.  One could easily do species ID by selecting a highly
 polymorphic locus such as beta-casein and sequencing the amplified
 product.

 TTGATTGCTAAACACTGGGC
 Eric E. Snyder
 Department of MCD Biology            We are not suspicious enough
 University of Colorado, Boulder      of words, and calamity strikes.
 Boulder, Colorado 80309-0347
 LeuIleAlaLysHisTrpAl



  Article 1328 of sci.chem:
 From: Keith Conover < ireland@ac.dal.ca >
 Date: 27 Nov 90
 Organisation: Dalhousie University, Halifax, Nova Scotia, Canada

 In article &lt;...>, andrewt@cs.su.oz (Andrew Taylor) writes:
 > In article &lt;...> roy@phri.nyu.edu (Roy Smith)
 >>...
 >> to remember reading something about doing PCR on DNA from a frozen
 > quagga Quaggas became extinct about 100 years ago (in Southern
 > Africa). Maybe you have the wrong animal?

 If I remember correctly, two different experiments are being confused
 here. The quagga DNA came from a museum specimen, possibly a hide. The
 PCR done on DNA isolated from a beast which was frozen and extinct from
 the last ice age came from a wooly mammoth.

 Keith Conover
 


 Article 1337 of sci.chem:
 From: David Throop <throop@cs.utexas.edu>
 Date: 29 Nov 90
 Organization: Dept of Computer Sciences, UTexas, Austin

   Walter Henry is trying to validate a claim that a particular document
 was signed in blood 300 yrs ago.
   Walter, can you be a little more specific?  Are you trying to
 determine-
   - That the signature is really in blood?
   - That the signature is really in human blood?
   - That the signature is really in the human blood of the putative
 author?
   - That the signature is really 300 yrs old?
   - That the signature is really in the handwriting of the putative
 author?
   - That the signature in blood is legally binding for the sale of a
 human soul?

   Really, Walter, if you're going to have us help on this, you should
 at least tell us the good parts.  What kind of a document gets signed
 in blood?  Or did the guy's quill pen just happen to be dull that day?
 This has got to be juicy.

   I'll assume that you are trying to prove that the signature is really
 in blood, but that you'd be willing to risk being fooled by other
 mammalian blood, and that you'll fix the document's age by other means.

   The methods advanced so far [various x-ray methods to detect iron and
 PCA reactions on the DNA] seem inadequate - in particular, I doubt
 you'll get any kind of a reading of the x-ray because of gross
 contamination, and, as others have pointed out, you may well not get
 any DNA from blood.

   I'd suggest gas chromatography/mass spec.  I remember articles circa
 1979 about using GCMS to identify different bacterial strains - they'd
 culture the bacteria, take a couple of micrograms, inject the sample
 into a very hot chamber for flash pyrolysis, and then routed the
 pyrolysis products to the gc column, then ms.  The pyrolysis signatures
 had enough detail (and were repeatable enough) that they could tell
 different strains of the same organism apart, but still identify a
 broad range of pathogens.

 And of course, GCMS is much more forgiving of contamination than other
 methods, because if you see the same GC peaks as the pure sample, and
 those peaks have the same MS signatures as the pure sample, you've got
 a positive no matter what other garbage comes out in the other peaks.
 And GCMS works with extremely small sample sizes.

 I'd think that this approach would certainly be good enough to pick out
 blood from non-blood, and probably human from non-human blood.

 But finally, I'd say check in with a forensic pathologist - crime lab
 folks.  They work with identifying blood from contaminated samples all
 the time.  They've probably got lots of techniques that us general
 chemistry types wouldn't know about.

 David Throop



 Article 1378 of sci.chem:
 From: Larry Lippman <larry@kitty.UUCP>
 Summary: Analytical methods employing pyrolysis with GCMS
 Date: 5 Dec 90

 In article <...>, throop@cs.utexas.edu (David Throop) writes:
 >  I'd suggest gas chromatography/mass spec.  I remember articles circa
 > 1979 about using GCMS to identify different bacterial strains - they'd
 > culture the bacteria, take a couple of micrograms, inject the sample
 > into a very hot chamber for flash PYROLYSIS, and then routed the
 > PYROLYSIS products to the gc column, then ms.  The PYROLYSIS
 > signatures had enough detail (and were repeatable enough) that they
 > could tell different strains of the same organism apart, but still
 > identify a broad range of pathogens.

     I do not believe that PYROLYSIS followed by GCMS will be of any use
 for identification of blood.  PYROLYSIS has its place for the
 identification of functional groups through the creation of smaller
 molecules such as: CH4, C2H4, C2H6, CO, CO2, NH3, H2S, H2O, etc. The
 identification of blood is based upon the characteristics of a
 *structure* - hemoglobin.  Don't forget that as a complex protein,
 hemoglobin has an m.w. of around 65,000.  Mere functional group
 identification analysis (in this case not that much better than simple
 CHNO determination, for that matter) is not likely to create an a
 priore basis to differentiate hemoglobin from say, myoglobin, serum
 albumin, or other organic materials.

     PYROLYSIS followed by GCMS is a useful technique in forensic
 science, but not for identification of blood or other body fluids.
 PYROLYSIS followed by GCMS is useful for rapid toxicology screening
 (where one has *real* unknowns!), especially when such GCMS results are
 compared to standards by means of Kovats retention indices (much work
 has been done along such lines).

 > And of course, GCMS is much more forgiving of contamination than other
 > methods, because if you see the same GC peaks as the pure sample, and
 > those peaks have the same MS signatures as the pure sample, you've got
 > a positive no matter what other garbage comes out in the other peaks.
 > And GCMS works with extremely small sample sizes.

     The above is generally true.

 > I'd think that this approach would certainly be good enough to pick
 > out blood from non-blood, and probably human from non-human blood.

     If we had a *singular* unknown and compared a standard sample of
 blood to an unknown sample which bore no relation to blood (such as
 ink), I would agree.  However, there are too many unknown factors due
 to the age of the document, its material and potential impurities
 (paper, parchment, coatings, etc.) to be confident of any such
 identification.

     With respect to differentiation of "human from non-human blood", I
 don't believe there is a prayer of a chance that such identification
 can be made using GCMS - provided that we are comparing human to other
 mammalian blood.

     PYROLYSIS followed by GCMS has an outstanding application: the
 identification of polymers, especially mixed polymers.

 Larry Lippman 

***********************************************************************